Another way to make an opto-compressor

It's just the basic idea I want to throw out there, to start a discussion; if it has any merit at all there would be plenty of ways to implement it.  I'm going to miss-use terms here, I know it....  The idea takes a balanced topography and sums the two signals to attenuate the audio, instead of using an LDR as a voltage divider.  Umm--  my imagination and knowledge limits me to the vision of a green mic-pre type front end, only instead of a gain knob there is an optocoupler (possibly in parallel with a resistor to bring it to the right range of values?).
So what do you all think?  :

that is cool. I wonder how summing the 2 anti-phase signals differs from shunting ( a SE circuit) to ground. would the attenuation amount/response be different between the 2 methods with all other things constant?

From a quick review it seems like six of one half dozen another. I guess you could tweak values to null to give more attenuation, but would perhaps trade away some noise floor and add complexity. Distortion should be a wash if executed well.  Go for it...

JR

> sums the two signals to attenuate the audio

Not any "bad" idea. The original patent has several "bridge" networks to do exactly that.

You can reach an infinite null, but it is infinitely fiddly to stay on-null.

How bad do you need to abuse your music? The simple divider can give 30dB GR at 100dB S/N or 50dB GR at 90dB S/N.